Fractional lithium clearance (CLi/CIn), transit time occlusion time (e-TT/OT), and late proximal tubule fluid-to-plasma inulin ratio [1/(TF/P)In] determined by micropuncture were measured in the same kidney during 5-sec-butyl-5-ethyl-2-thiobarbituric acid (Inactin) anesthesia in rats fed either less than 2 mu eq sodium/g diet, rodent chow containing 20 mu eq sodium/g diet, or rodent chow and 0.5% NaCl to drink. Both CLi/CIn vs. e-TT/OT (r = 0.59) or 1(TF/P)In (r = 0.65) were linearly correlated at all sodium intakes. Quantitative estimates of proximal delivery by 1/(TF/P)In and e-TT/OT were similar (P = NS), whereas estimates from CLi/CIn were lower (P less than 0.05 or less) than the direct methods even when corrected for pars recta reabsorption. After acetazolamide or amiloride, but not furosemide, all methods for proximal delivery correlated quantitatively (P = NS) in sodium-restricted rats. Between amiloride and nondiuretic sodium-restricted rats, 1/(TF/P)In was not different. Proximal delivery estimated by CLi/CIn slightly (P = NS) underestimated direct measurements during 2% volume expansion [fractional sodium excretion (FeNa) 0.65 +/- 0.08%]. Thus, in the rat, CLi/CIn is not a quantitative estimate of proximal delivery if FeNa is less than 0.65% due to distal lithium reabsorption. CLi/CIn determined after amiloride may provide a correct estimate of proximal delivery during sodium restriction.